Abstract: When solving multimodal optimization problems, the lion swarm optimization (LSO) algorithm will face many problems, such as low individual diversity, slow search speed, premature convergence or even falling into local extremum. The traditional approach is to introduce chaotic search, gaussian mutation or other mutation strategies to enhance the local search ability of the LSO algorithm. These improvements have been verified and the performance of the algorithm can be improved to a certain extent. However, these improved strategies lack effective use of population information, which will still affect the performance of the algorithm in search speed and search accuracy. To solve this problem, on the basis of the above-mentioned improved algorithm, this paper introduces the distribution estimation algorithm and proposes a hybrid improved LSO algorithm. The hybrid improved algorithm analyzes and learns the structure of the problem by constructing a probability model for the dominant group, and guides the efficient optimization of individuals in the population according to this information. It is verified in five standard test functions and ten test functions of IEEE CEC 2021. Compared with the traditional improved LSO algorithm, the results show that the hybrid improved LSO algorithm is superior.
原文链接
分享
引用
